Experimental Infections with Mycoplasma agalactiaeIdentify Key Factors Involved in Host-ColonizationEric Baranowski1,2, Dominique Bergonier2,1, Eveline Sagne2,1, Marie-Claude Hygonenq1,2,

Patricia Ronsin2, Xavier Berthelot2,1, Christine Citti1,2*

1 INRA, UMR 1225, IHAP, Toulouse, France, 2 Universite de Toulouse, INP-ENVT, UMR 1225, IHAP, Toulouse, France

Abstract

Mechanisms underlying pathogenic processes in mycoplasma infections are poorly understood, mainly because of limitedsequence similarities with classical, bacterial virulence factors. Recently, large-scale transposon mutagenesis in the ruminantpathogen Mycoplasma agalactiae identified the NIF locus, including nifS and nifU, as essential for mycoplasma growth in cellculture, while dispensable in axenic media. To evaluate the importance of this locus in vivo, the infectivity of two knock-outmutants was tested upon experimental infection in the natural host. In this model, the parental PG2 strain was able toestablish a systemic infection in lactating ewes, colonizing various body sites such as lymph nodes and the mammary gland,even when inoculated at low doses. In these PG2-infected ewes, we observed over the course of infection (i) thedevelopment of a specific antibody response and (ii) dynamic changes in expression of M. agalactiae surface variableproteins (Vpma), with multiple Vpma profiles co-existing in the same animal. In contrast and despite a sensitive model, noneof the knock-out mutants were able to survive and colonize the host. The extreme avirulent phenotype of the two mutantswas further supported by the absence of an IgG response in inoculated animals. The exact role of the NIF locus remains tobe elucidated but these data demonstrate that it plays a key role in the infectious process of M. agalactiae and most likely ofother pathogenic mycoplasma species as many carry closely related homologs.

Citation: Baranowski E, Bergonier D, Sagne E, Hygonenq M-C, Ronsin P, et al. (2014) Experimental Infections with Mycoplasma agalactiae Identify Key FactorsInvolved in Host-Colonization. PLoS ONE 9(4): e93970. doi:10.1371/journal.pone.0093970

Editor: Mitchell F. Balish, Miami University, United States of America

Received January 16, 2014; Accepted March 10, 2014; Published April 3, 2014

Copyright: � 2014 Baranowski et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This work was supported by grants from INRA (Animal Health Division, AAP 2012) and ENVT. The funders had no role in study design, data collectionand analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declared that no competing interests exist.

* E-mail: c.citti@envt.fr

Introduction

The term ‘‘mycoplasma’’ designates a large group of bacteria

whose common feature is the lack of a cell wall. These atypical

organisms belong to the class Mollicutes that has evolved by massive

losses of genetic material from Gram-positive ancestors with low

G+C content [1–3]. Successive genome reductions have left

current mycoplasmas with an amount of genetic information that

is close to the minimal requirements for sustaining autonomous life

[4,5], explaining in part their parasitic life-style. This strict

association is often damaging to the host and several mycoplasma

species are successful pathogens, capable of establishing persistent

infections and causing debilitating diseases in humans and a wide

range of animals [1]. Since most mycoplasmas are able to replicate

in axenic media, this particular group of bacteria includes some of

the smallest and simplest life-forms known. Along with the concept

of a minimal cell, they have offered an experimental platform to

explore and create the first organism controlled by a chemically

synthesized genome [6]. While these studies have greatly

expanded our understanding of cell biology fundamentals, factors

involved in mycoplasma-host interactions remain poorly under-

stood.

Of the mycoplasmas species relevant to the veterinary field,

Mycoplasma agalactiae is an attractive model to study the infectious

process of mycoplasmas in the natural host. First, it is an important

pathogen responsible for contagious agalactia (CA) in small

ruminants, an economically important disease notifiable to the

World Organization for Animal Health [7–9]. Secondly, recent

genomic analysis revealed that large portion of its genome has

undergone gene exchanges with members of the mycoides cluster

[10]. Although phylogenetically distant from M. agalactiae, all

members of this cluster are ruminant pathogens, several of which

being also responsible of CA in small ruminants. Among the genes

that were exchanged, several encode surface and membrane

components that may contribute to the infectious process. Finally,

an additional interest in studying M. agalactiae is its close

phylogenetic and genetic proximity with M. bovis [10], a re-

emerging pathogen of growing concern for the cattle industry that

induces similar clinical manifestations, including mastitis, arthritis

and pneumonia [11]. Both ruminant pathogens possess a family of

related surface proteins which high-frequency variation in

expression and structure modulate the mycoplasma surface

composition and accessibility, providing a means for these

organisms to rapidly adapt to various environmental changes.

Phase and size variations have been shown to occur in vitro with a

frequency estimated at 1022 to 1025 events/cell/generation [12].

The type strain of M. agalactiae, strain PG2, carries a simplified

version of this family, which members are designated as Vpmas

(Variable protein of M. agalactiae) and are encoded by six genes,

vpma U to Z, occurring as a cluster on the chromosome. In a given

cell, only one vpma gene is expressed while in the overall

population the six products occur at different rates [10,12,13].

PLOS ONE | www.plosone.org 1 April 2014 | Volume 9 | Issue 4 | e93970

This intra-clonal variation is due to the presence of a unique

promoter at the vpma locus that is alternatively placed via a cut and

paste mechanism in front of silent vpma genes, resulting in ON and

OFF switching in Vpma expression [12]. This phenomenon is

known to occur in vitro although it has never been truly

documented over the course of an infection, despite its potential

role in host colonization and in escaping the host-immune system.

Whole genome sequencing of a number of mycoplasmas species

failed to reveal similarities with virulence factors identified in more

classical bacteria suggesting that these minimal pathogens have

evolved unconventional strategies to interact with their hosts [1].

An exception includes a magnesium-dependent nuclease homo-

logue to the staphylococcal SNase, which role in M. agalactiae host-

survival or host-pathogenicity has still to be demonstrated [14].

New opportunities to investigate factors involved in mycoplasma-

host interactions have emerged through the development of

molecular tools for the genetic manipulation of mycoplasmas [15–

17]. Recently, we reported the development of a cell culture assay

for the high-throughput screening of mutants with reduced growth

capacities upon co-culture with host cells [18,19]. Using this

approach, 62 genomic loci of M. agalactiae were identified out of a

library of ca. 2000 mutants as contributing to this phenotype. Of

the 62 loci potentially involved in M. agalactiae host-interaction,

one was of particular interest because two mutants having inserted

a transposon at different position of the same gene, nifS, displayed

most extreme growth-deficient phenotype in cell culture [18]. This

locus is composed of two genes, nifS and nifU that are co-

transcribed [19] and encodes homologues of SufS and SufU, two

proteins presumably involved in Fe-S cluster biosynthesis in Gram-

positive bacteria [20].

While cell culture provides a simple and efficient screening

system for genome-scale analysis of M. agalactiae loci involved in

host-interaction, the expression and regulation of the correspond-

ing biological functions are expected to be far more complex in the

host context, raising the question of the in vivo relevance of pre-

screening cell assays. In this study, we formally addressed this issue

by defining the virulence of the NifS mutants in the natural host.

The in vivo model we developed consisted in inoculating lactating

ewes by the subcutaneous route with low doses of M. agalactiae and

to monitor mycoplasmas shedding in milk over three weeks of

infection. This model allowed us to evaluate the capacity of these

mutants to spread from the point of inoculation to the mammary

gland, and hence to address their ability to colonize, replicate and

survive in the animal host when compared to the wild type. We

also evaluated how the Vpma variations can contribute to the

success of the infection and possibly influenced the outcomes of

in vivo studies when using defined mutants. This was performed by

following the dynamic of the Vpma expression over the course of

infection. For this purpose, we monitored the Vpma phenotype of

organisms sequentially recovered in milk and the corresponding

humoral host response.

The inability of the two NifS mutants to generate a successful

infection in vivo indicates that key functions contributing to M.

agalactiae survival and dissemination were affected and will be

discussed. This study also brings into a new light the dynamic

changes in mycoplasma surface composition that occurs over the

course of the infection by showing the rapid in-host oscillation in

M. agalactiae Vpma expression, with multiple profiles co-existing in

the same animal.

Materials and Methods

Bacterial Strains and Culture ConditionsM. agalactiae reference strain PG2 was used in this study [10].

Mutants NifS1 and NifS2, originally described as 7.82 and 7.134,

were selected from a library of transposon knock-out mutants

generated in strain PG2 [18]. They both carry a transposon in a

genomic region designated as the NIF locus (CDS MAG0720 to

MAG0730) (Fig. 1). The reference strain PG2 and the NifS

mutants were cultured at 37uC in a modified SP4 medium, termed

as SP4-HS, in which horse serum (HS) was substituted for fetal

bovine serum. The SP4-HS medium was supplemented with

500 mg/ml cefalexin (Virbac). Mycoplasma cultures were stored at

280uC, and CFU titers were determined by serial dilutions in

Dubelco’s phosphate-buffered saline (DPBS; Invitrogen) supple-

mented with 1% heat inactivated HS (Invitrogen). Dilutions were

spotted (10 ml) onto solid SP4-HS medium and mycoplasma

colonies were counted after 2 to 5 days incubation at 37uC. The

absence of contaminating wild-type sequences in culture stocks of

NifS mutants was tested by PCR amplification, as previously

described [18].

Ethics StatementExperimentations were conducted as prescribed by the guide-

lines of the EU Council on Animal Care (86/609/CEE) and

approved by the ethics committee of Midi-Pyrenees (France) under

the agreement number MP/02/49/09/11. Animals were eutha-

nized at day 28 post-inoculation by an intravenous injection of

pentobarbital (140 mg/kg) (EUTHASOL VET, LE VET, Nether-

lands).

Animals and Experimental InfectionLactating ewes of the Lacaune breed, originating from a CA

free area located in the south-west of France (Aveyron), were used

for experimentations. Lactating ewes and their progeny were

allocated into five experimental groups (5 ewes per group) and

housed in separate rooms at animal facilities of the ENVT.

Animals were shown to be free of M. agalactiae and other ruminant

mycoplasma species, and exhibited undetectable levels of specific

antibodies against M. agalactiae. Serological analyses were per-

formed by ELISA (IDEXX M. agalactiae Screening Ab Test) and

Western blot analysis (see below).

For inoculum preparations, M. agalactiae reference strain PG2

and NifS mutants were grown in SP4-HS medium. After 48 hours

at 37uC, mycoplasma cultures were stored at 280uC as 1 ml

aliquots. Before inoculation, frozen aliquots of mycoplasma

cultures were thawed and CFUs were enumerated. On inoculation

day, aliquots from the same batch were thawed and adjusted to the

required CFU concentration by dilution in SP4-HS medium.

Samples were kept on ice before inoculation. Each animal received

a 2 ml subcutaneous injection of the inoculum or SP4-HS medium

at the left shoulder. The doses of viable mycoplasmas injected to

each experimental group were determined by titration of the

inoculum before and immediately after animal inoculation. PCR

amplifications failed to detect wild-type sequences in cultures of

the NifS mutants, ruling out any possible contamination with

either the parental strain or other mutants (data not shown).

Clinical Examination and Sample CollectionAnimals were examined and clinical signs were recorded and

scored from day 7 before inoculation to day 28 post-inoculation

(PI), with two- or three-days intervals. Milk and whole-blood

samples were collected at regular interval throughout the

experiments. Milk production was determined twice a week.

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 2 April 2014 | Volume 9 | Issue 4 | e93970

Figure 1. Genetic and antigenic features of M. agalactiae strain PG2 and NifS mutants used for animal inoculation. Knock-out mutantsNifS1 and NifS2 were generated by transposon mutagenesis in M. agalactiae reference strain PG2 [18]. (A) Drawing illustrating the NIF locus (greyarrows) in M. agalactiae. Genes are represented according to their orientations on the genome. Transposon insertion sites in NifS mutants areindicated by a vertical bar with a closed circle, an arrow indicate its orientation. The insertion of the transposon in mutant NifS2 occurred within thestop codon of nifS [18]. The scale is indicated. (B) SDS-PAGE analysis of Triton X-114 soluble materials extracted from cultures of PG2 and NifS mutantsfollowed by Coomassie staining or immunodetection using the PAL97 hyperimmune serum. Molecular weight standards (MW) are in kDa. (C)Characterization of phase variable, Vpma expression profile in M. agalactiae inoculums. Vpma lipoproteins U, V, W, X, Y and Z were detected using

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 3 April 2014 | Volume 9 | Issue 4 | e93970

Lymph nodes were examined and collected at necropsy (day 28),

and stored at 280uC.

Bacteriological and PCR Examination of BiologicalSamples

After collection, milk and whole-blood samples (collected in

EDTA tubes) were diluted 1:10 and 1:100 in 2 ml SP4-HS

medium (containing 500 mg/ml cefalexin and 0.01% thallium

acetate). Three days following incubation at 37uC, cultures were

distributed in 96-well plates (Falcon) and spotted onto solid SP4-

HS medium (containing 500 mg/ml cefalexin and 0.01% thallium

acetate) using a 96-pin replicator (Boekel Scientific). The volume

of the sample transferred by one pin of the replicator was

estimated at about 1 ml. The development of mycoplasma colonies

was observed after 7 to 14 days incubation at 37uC. The minimal

amount of detectable CFU was estimated at 102 CFU (56102

CFU/ml) by testing serial dilutions of a mycoplasma culture stock.

Lymph nodes were separated from the surrounding fat,

homogenized manually using sterile blades and resuspended in

2 ml SP4-HS medium. After 4 h incubation at 37uC, supernatants

were diluted 1:1 and 1:20 in 2 ml SP4-HS medium, and the

remaining suspensions were stored at 280uC. After 3 days

incubation at 37uC, primary cultures were sub-cultured under the

same conditions (dilution 1:10). Primary cultures and sub-cultures

were tested for the presence of mycoplasmas as described above.

Mycoplasma cultures were further examined by PCR amplifi-

cation. Samples (1.5 ml) were centrifuged at 12,0006g for 20 min.

The bacterial pellet was resuspended in 50 ml lysis buffer (0.1 M

Tris-HCl, pH 8.5; 0.05% Tween 20; 0.25 mg/ml proteinase K),

incubated at 37uC for 1 h followed by proteinase K inactivation at

95uC for 10 min. PCR amplifications (50 ml) were performed as

described previously [21] using 2 ml of lysates and the M. agalactiae-

specific primers MAPol-1F (59-CATTGAACCTCTTATGT-

CATTTACTTTG-39) and MAPol-5R_mod (59-CTATGTCAT-

CAG-CTTTTGAGTGA-39), which span genomic positions

78626 to 78653, and 78866 to 78888, respectively.

SDS-PAGE and Western Blot Analysis of M. agalactiaeTriton X-114 Extracts

Mycoplasmas grown in SP4-HS medium were collected by

centrifugation at 10,0006g and resuspended in DPBS (Invitrogen).

Protein concentration was determined using the Quick Start

Bradford protein assay (Bio-Rad). Triton-X114 soluble proteins

were extracted with 1/9 volume of Triton X-114 solution (10 mM

Tris-HCl pH 7.4; 150 mM NaCl; 10% Triton X-114). After 1 h

incubation at 4uC under mild agitation, aqueous and detergent

phases were separated by incubation at 37uC. After 5 min

centrifugation at 12,0006g, the detergent phase was found as an

oily droplet at the bottom of the tube. The detergent phase

containing most hydrophobic, membrane proteins was washed 3

times with 9 volumes of Tris-buffered saline (TBS) (10 mM Tris-

HCl pH 7.4; 150 mM NaCl), 1 h incubation at 4uC and phase

separation at 37uC. Triton X-114 was removed by protein

precipitation in methanol. Proteins resuspended in Laemmli

sample buffer were separated by SDS-PAGE electrophoresis using

the Mini-Protean electrophoresis system (Bio-Rad). Proteins were

visualized by staining using the Bio-Safe Coomassie stain (BioRad).

For Western blotting, proteins were transferred to Protran

nitrocellulose membranes (Whatman). The amount of Triton X-

114 soluble material transferred to membrane strips was equiv-

alent to 108 CFU of M. agalactiae. Membranes were blocked in

TBS containing 5% skimmed-milk for 2 hours, then incubated

overnight at 4uC with experimental sera diluted (1/125) in TBS

containing 0.05% Tween 20 and 10% decomplemented horse

serum (Invitrogen) or 1% skimmed milk. Western blots were

developed with an HRP-labeled secondary antibody (DAKO;

P0163) and 4-chloro-naphtol as substrate. The PAL97 hyperim-

mune serum was collected from a sheep naturally infected by M.

agalactiae. The anti-P80 sheep serum, kindly provided by P.

Giammarinaro, was produced by animal immunization with a P80

recombinant protein (data not shown). Vpma-specific polyclonal

antibodies were used as previously described [22].

Results

Subcutaneous Inoculation with Low Doses of M.agalactiae: Colonization of the Udder and Sheddingin Milk

Experimental models for ruminant mycoplasmoses are generally

conducted with high infectious doses inoculated via the respiratory

tract, the eye conjunctiva or the mammary gland. These

parameters were not best suited for our study as we aimed at (i)

monitoring the mycoplasma systemic colonization of the mam-

mary gland from a different point of entry, (ii) recovering

mycoplasmas during the course of infection without invasive

procedures, and (iii) controlling the size of the inocula injected de

facto. As well, we wanted to define whether low infectious dose can

be used for future signature-tagged mutagenesis approaches, not

described here.

Two experimental groups of lactating ewes were inoculated

subcutaneously with 103 or 105 CFU of PG2 (experimental groups

PG2-103 and PG2-105, respectively), while a control group

received growth medium alone by the same route (Table 1).

Animal behavior was monitored during 28 days and remained

unchanged throughout this period, without any decrease in

appetite or significant difference in rectal temperatures in between

experimental and control groups (data not shown). Milk produc-

tion remained stable in each group, with only two animals, animal

A5 in group PG2-103 and animal A8 in group PG2-105, exhibiting

an hypogalactia of the left udder-half (80% reduction in milk

production). No other clinical sign was observed in infected and

control animals. Subclinical mammary infections with Escherichia

coli or coagulase-negative Staphylococci were identified in groups

PG2-103 (animals A4 and A5) and PG2-105 (animals A6 and A8)

(Table 1).

The capacity of M. agalactiae to infect its natural host and to

colonize the mammary gland was investigated in milk samples

collected at regular intervals from up to 28 days after inoculation.

As shown in Table 2, all experimentally infected animals (except

A7, see below) were found to excrete mycoplasmas in milk, while

no mycoplasma was recovered from the control group. One

exception was animal A7 in group PG2-105, which remained

negative for all samples tested. Two patterns of excretion clearly

emerged in the experimental groups infected by PG2. One

concerned a majority (8/10) of animals and was characterized by a

transient excretion between days 18 and 28 post-inoculation

(Table 2). The other was limited to one animal in each group, A5

specific polyclonal antibodies (see Materials and Methods). A lipoprotein P80 antiserum was used as control. (D) Growth curves of PG2 and NifSmutants in SP4-HS medium. Mycoplasma titers used for inoculations were 107 CFU/ml (left panel) and 105 CFU/ml (right panel). The data are themeans of three independent assays. Standard deviations are indicated by error bars.doi:10.1371/journal.pone.0093970.g001

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 4 April 2014 | Volume 9 | Issue 4 | e93970

in group PG2-103 and A8 in group PG2-105, and corresponded to

a continuous excretion that started as early as 7 days post-

inoculation and lasted until the end of the experiment. Mycoplas-

ma excretion was mainly detected in a single udder-half in 21/27

positive samples. One exception was animal A4 in group PG2-103,

for which mycoplasmas were always isolated from both udder-

halves (Table 2). Examination of whole blood samples collected at

different times throughout the experiment failed to reveal the

presence of mycoplasmas, with a detection limit estimated at

102 CFU (56102 CFU/ml). Negative cultures of whole blood and

milk samples were further analyzed by PCR amplification that

confirmed the absence of mycoplasmas in these samples (data not

shown). The presence of M. agalactiae in positive cultures was

confirmed by PCR amplification using species-specific primers

(data not shown).

The dissemination and survival of M. agalactiae in lactating ewes

was further investigated by examination of lymph node samples

collected at necropsy (Fig. 2A). Here, PCR amplification was

performed in parallel to cultivation in order to overcome the

outgrowth by contaminating bacteria or fungi that occurred

occasionally despite of the presence of cefalexin and thallium

acetate in the medium. With the exception of three animals (A2,

A3 and A6), all infected ewes displayed at least one positive lymph

node, with a total of 80% of the samples (10/12) being PCR

positive (Fig. 2A). Lymph nodes frequently found positive were the

left prescapular, next to the site of inoculation (left shoulder) and

the supramammary lymph nodes, (see Fig. 2A and Table 2).

Nevertheless, the PCR positive iliac lymph node detected in

animal A8 suggests a broader dissemination for M. agalactiae.

These data indicated that M. agalactiae is able to survive within the

host even for long periods, and most likely in the lymph nodes.

Altogether, these results indicate that high infectious doses are

not required for successful infection and dissemination of M.

agalactiae in the natural, ovine host. Thus, the experimental

infection of lactating ewes by subcutaneous route and the

subsequent recovery of mycoplasmas from milk samples provide

a simple and non-invasive assay for in vivo screening of large pools

of M. agalactiae mutants.

Serological Response and Dynamics of Variable VpmaSurface Proteins

The development of a serological response in animals infected

with M. agalactiae was monitored by Western blotting (Table 2).

Despite a 100-fold difference in the dose used for animal

inoculation, both experimental groups developed IgG antibodies

to M. agalactiae surface antigens. Yet, the appearance of specific

IgG in group PG2-103 was delayed when compared to group

PG2-105.

To better characterize the development of the humoral response

within single animals, sera from animals A1, A5, A8 and A9,

which exhibited a strong reactivity against M. agalactiae, were used.

Sera collected at regular intervals throughout the experiment were

analyzed by Western blotting using the Triton-X114 soluble

proteins of strain PG2 that contain most hydrophobic, membrane

proteins (Fig. 3). Immunoprofiles revealed differences among

animals in terms of (i) the kinetic of the IgG response, with animals

A1, A8 and A9 having a strong IgG response from day 11 while

that in A5 appeared to be delayed and (ii) specificity, with the 4

animals displaying different immunoprofiles. A strong reactivity

was mainly obtained with proteins migrating between 35 and

55 kDa (Fig. 3). More specifically, antibodies reacted strongly with

two protein bands showing molecular weight values similar to

Vpmas V (44 kDa) and Y (39 kDa), two phase variable

lipoproteins of M. agalactiae (data not shown).

Differences in the development of the humoral response to M.

agalactiae prompted us to compare the antigenic profile of M.

agalactiae isolated from milk samples. For this purpose, the Triton-

X114 soluble proteins from seven mycoplasma-positive cultures,

corresponding to animals positive at day 28 (A4, A5 and A8) or to

animals that were positive once (A1, A2, A6 and A9), were

analyzed by SDS-PAGE and Western blotting. As shown in

Figure 4, their profile revealed major differences in protein

expression. More specifically, these were associated to highly

abundant proteins with molecular weight values that ranged

between 35 and 55 kDa and that were predominantly recognized

by the PAL97 hyperimmune serum (Fig. 4). These data suggested

that M. agalactiae isolated from milk samples were expressing

different Vpma lipoprotein profile and this was confirmed by using

Vpma-specific polyclonal antibodies (Fig. 4).

While the PG2 inoculum expressed predominantly Vpma Y

(Fig. 1), the Vpma profile exhibited by the seven isolates was highly

heterogeneous, with some samples expressing predominantly

Vpma W (A5), Vpma U (A2), Vpma V and Z (A4), or a complex

mixture of several Vpmas reflecting the presence of several

populations in the sample. A lipoprotein P80 antiserum was used

as control and confirmed the constant expression of this

lipoprotein in all the seven isolates. In contrast to M. agalactiae

milk isolates, the Vpma expression profile exhibited by PG2 and

Table 1. Overview of the experimental infections with M. agalactiae strain PG2 and NifS mutants in lactating ewes using thesubcutaneous route.

Animalgroup

Infectiousdose (CFUs)a

Hypo-galactiab

Milksamplesc

Serologicalresponsed

Lymphnodese

Otherbacteriaf

Control – 0/5 0/5 0/5 0/5 2/5

PG2-103 3.261.06103 1/5 5/5 5/5 3/5 2/5

PG2-105 2.360.26105 1/5 4/5 5/5 4/5 2/5

NIF1 1.660.26105 0/5 0/5 0/5 0/5 0/5

NIF2 1.660.36105 0/5 0/5 0/5 0/5 0/5

aMean mycoplasma CFUs inoculated to each experimental group;bnumber of animals with reduced milk production;cnumber of animals with at least one milk sample found positive for M. agalactiae;dnumber of animals that developed a serological response to M. agalactiae;enumber of animals with at least one lymph node found positive for M. agalactiae (the lymph nodes analyzed are indicated in Fig. 2);fnumber of animals with subclinical mammary infections with Escherichia coli or coagulase-negative Staphylococci.doi:10.1371/journal.pone.0093970.t001

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 5 April 2014 | Volume 9 | Issue 4 | e93970

Ta

ble

2.

Myc

op

lasm

ae

xcre

tio

nan

dse

rolo

gic

alre

spo

nse

inan

imal

sin

ocu

late

dw

ith

M.

ag

ala

ctia

est

rain

PG

2an

dN

ifS

mu

tan

ts.

Bio

log

ica

la

ssa

yA

nim

al

gro

up

An

ima

lD

ay

po

stin

ocu

lati

on

0–

47

91

11

41

61

82

12

32

52

8

A1

22

22

22

22

R2

2

A2

22

22

22

2R

22

2

PG

2-1

03

A3

22

22

22

2L

22

2

A4

22

22

22

22

LRLR

LR

A5

22

2L

LL

LL

LRL

LR

Iso

lati

on

of

M.

A6

22

22

22

L2

22

2

ag

ala

ctia

ein

A7

22

22

22

22

22

2

milk

sam

ple

saP

G2

-10

5A

82

LL

LL

LRL

RL

LL

A9

22

22

22

22

2L

2

A1

02

22

22

22

22

L2

NIF

1A

11

toA

15

22

22

22

22

22

2

NIF

2A

16

toA

20

22

22

22

22

22

2

A1

22

2+

++

++

++

+

A2

22

22

22

2+

++

+

PG

2-1

03

A3

22

22

++

++

++

+

A4

22

22

22

22

22

+

A5

22

22

22

2+

++

+

Sero

log

ical

IgG

A6

22

22

++

++

++

+

resp

on

seb

A7

22

22

++

++

++

+

PG

2-1

05

A8

22

++

++

++

++

+

A9

22

++

++

++

++

+

A1

02

22

2+

++

++

++

NIF

1A

11

toA

15

22

22

22

22

22

2

NIF

2A

16

toA

20

22

22

22

22

22

2

aT

he

exc

reti

on

of

M.

ag

ala

ctia

ein

the

mam

mar

yg

lan

dis

ind

icat

ed

wit

ha

lett

er

ind

icat

ing

ud

de

rh

alve

sfo

un

dp

osi

tive

:ri

gh

t(R

),le

ft(L

)o

rb

oth

(LR

);b

the

de

tect

ion

of

spe

cifi

cIg

Gan

tib

od

ies

toT

rito

nX

-11

4so

lub

lean

tig

en

so

fM

.a

ga

lact

iae

inan

imal

sera

are

ind

icat

ed

(+).

do

i:10

.13

71

/jo

urn

al.p

on

e.0

09

39

70

.t0

02

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 6 April 2014 | Volume 9 | Issue 4 | e93970

NifS cultures used for animal inoculations were homogeneous

(compare Vpma profiles in Fig. 1 and Fig. 4), apart from the

inherent background in alternate Vpma populations and despite

differences in their number of passages under laboratory

conditions.

The long term excretion of mycoplasmas in the left half-udder

of animals A5 and A8 allowed us to analyze changes in M.

agalactiae Vpma profile over a period of 18 to 22 days. SDS-PAGE

analysis of Triton-X114 soluble proteins from mycoplasma

cultures indicated continuous changes in protein expression, and

Vpma-specific polyclonal antibodies revealed a highly dynamic

Vpma profile with stochastic changes in Vpma expression (Fig. 5).

To further investigate changes in the expression of M. agalactiae

surface proteins upon replication in vivo, Triton-X114 soluble

proteins of mycoplasmas recovered from lymph nodes were

analyzed (Fig. 2B). As observed with milk isolates, important

differences in protein expression profile were also found with

mycoplasmas isolated from the axillar, prescapular and supra-

mammary lymph nodes, suggesting that mycoplasma diversifica-

tion is not restricted to the mammary gland, but can occur at

different stage of the infection.

At day 28, mycoplasmas that circulated in animals A4 and A5

differed in their protein expression depending on whether they

were isolated from the milk, or from the left- or right-

supramammary lymph nodes (Fig. 2C), suggesting that diversifi-

cation occurred independently in each environment.

Overall, these results indicate that (i) even a single passage in vivo

can severely affect the M. agalactiae surface composition, (ii) several

changes in Vpma expression may occur in the course of the

infection in a single animal, and (iii) different Vpma profiles can be

detected in different organs of the same animal.

M. agalactiae Knock-out Mutants Selected in Cell Cultureare Unable to Survive and Colonize the Ovine Host

Previous studies have identified the M. agalactiae NIF locus,

composed of the co-transcribed nifS and nifU genes, as essential for

Figure 2. Recovery of M. agalactiae from the lymph nodes of lactating ewes experimentally infected by the subcutaneous route. (A)Schematic drawing illustrating the recovery of M. agalactiae from animal lymph nodes in experimental groups PG2-103 and PG2-105. Experimentalgroups were inoculated subcutaneously at the left shoulder with different doses of M. agalactiae strain PG2 (see Table 1). The lymph nodes collectedat day-28 pi are indicated by an open circle (L-AXL: left-axillary; ILQ: Iliac; MED: mediastinal; MES: mesenteric; L-PSC: left-prescapular; L-SPM: left-supramammary; R-SPM: right-supramammary; RTP: retropharyngeal; PAR, parotid). For each group, lymph node samples found positives by PCRamplification are indicated (+) and the corresponding animals are given in parenthesis. M. agalactiae was isolated from all PCR positive samples,except those designated by an asterisk (*). (B) SDS-PAGE analysis of Triton X-114 soluble proteins of mycoplasmas recovered from animal lymphnodes in experimental groups PG2-103 (animals A4 and A5) and PG2-105 (animals A7, A8, A9 and A10). (C) Illustration of mycoplasma membraneprotein diversification upon PG2 (PG2) dissemination in the animal host. Mycoplasmas were isolated from (L) left- and (R) right- retromammary (RTM)lymph nodes or udder-halves (Udr). Proteins were visualized by Coomassie staining after SDS-PAGE. Molecular weight standards (MW) are in kDa.doi:10.1371/journal.pone.0093970.g002

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 7 April 2014 | Volume 9 | Issue 4 | e93970

Figure 3. Development of a serological response to M. agalactiae in lactating ewes experimentally infected by the subcutaneousroute. Immunoblotting pattern of M. agalactiae Triton X-114 soluble proteins obtained using sera from experimental groups PG2-103 (animals A1

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 8 April 2014 | Volume 9 | Issue 4 | e93970

mycoplasma growth in cell culture and suggested a possible role of

this locus in vivo [18]. To formally address this issue, two groups of

ewes, designated NIF1 and NIF2, were inoculated with 105 CFU

of the NifS1 and NifS2 mutants respectively (Table 1). Each of

these mutants derived from the PG2 type strain used above and

harbored a stable-transposon inserted at different positions in nifS

but exhibited identical growth-deficient phenotypes upon co-

cultivation with mammalian cells [18], while displaying wild-type

growth in axenic conditions (Fig. 1D). Complementation studies

previously confirmed the role of the NIF locus for M. agalactiae

survival in cell culture and restored the wild-type phenotype [18].

Yet, complemented strains were not used in our study to avoid the

use of a constant antibiotic treatment of the animals that would be

required to preventing the rapid loss of extra-chromosomal

plasmid carrying the wild-type genes. Instead, we hypothesized

that the use of the two mutants in two independent groups, if

resulting in the same outcome, would confirm the role of the NIF

locus in vivo.

Biological samples from experimental groups NIF1 and NIF2

were collected and examined as for PG2 groups. In contrast to the

two groups inoculated with PG2, attempts to isolate mycoplasmas

from the milk of animals of groups NIF1 and NIF2 failed. No

isolate was obtained from blood, as for PG2 groups. All animals

remained negative throughout the experiment with a detection

limit of our assay estimated at 102 CFU (56102 CFU/ml).

Negative cultures were further tested by PCR amplification that

confirmed the absence of mycoplasma material in the samples

tested. At necropsy, all lymph nodes tested were also found

negative for mycoplasma (Table 1).

The inability of the NifS mutants to survive and colonize the

host was further confirmed by the analysis of IgG antibody

response that remained undetectable in the serum of all inoculated

and A5) and PG2-105 (animals A8 and A9) collected at different days post-inoculation (0 to 28). The position at which lipoprotein P80 (80 kDa) andphase variable Vpma lipoproteins V (44 kDa) and Y (39 kDa) migrate are indicated. Molecular weight standards (MW) are in kDa.doi:10.1371/journal.pone.0093970.g003

Figure 4. Diversification of M. agalactiae antigenic structure following a single passage in the animal host. (A) SDS-PAGE analysis ofTriton X-114 soluble materials extracted from cultures of M. agalactiae recovered from milk samples at day 28 post-inoculation in experimentalgroups PG2-103 (animals A1, A2, A4 and A5) and PG2-105 (animals A6, A8 and A9) followed by Coomassie staining, immunodetection using the PAL97hyperimmune serum, and (B) characterization of the Vpma expression profile using specific polyclonal antibodies (see Materials and Methods). Alipoprotein P80 antiserum was used as control. Molecular weight standards (MW) are in kDa.doi:10.1371/journal.pone.0093970.g004

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 9 April 2014 | Volume 9 | Issue 4 | e93970

animals (Table 2). The antigenic profile of the two mutants was

closely similar to that of the parental strain PG2, when using the

PAL97 hyperimmune serum collected from a sheep naturally

infected by M. agalactiae (Fig. 1). This excluded the possibility of a

major change in the antigenic structure of the NifS mutants that

could account for the lack of IgG detection. This was further

confirmed by using Vpma-specific polyclonal antibodies that

identified Vpma Y as the main phase variable surface lipoprotein

expressed by each population, with subtle differences in Vpma

lipoproteins expressed at background levels (Fig. 1). Finally, the

absence of specific IgG antibodies in animals from groups NIF1

and NIF2 was further confirmed by Western blotting experiments

carried out using NifS1 and NifS2 surface antigens, instead of PG2

materials (data not shown).

These results clearly demonstrate the essential role played by

the NIF locus in M. agalactiae survival and colonization of the ovine

host. They also demonstrate that cell culture provide a useful

strategy for the pre-screening of large libraries of M. agalactiae

mutants.

Discussion

The avirulent phenotype exhibited by two knock-out mutants

harboring a transposon inserted at different positions of the single-

copy NifS gene (MAG0720) indicate that a key virulence factor of

M. agalactiae maps within the NIF locus (MAG0720 and

MAG0730). In contrast to the parent PG2 strain, M. agalactiae

NIF mutants were unable to establish an infection in the ovine host

as no mycoplasma nor specific antibodies were detected over the

course of the experiment in either of the two groups infected by

one or the other mutants. It is thus unlikely that other genetic

alterations could be responsible for the avirulence of the two NifS

mutants, since these were identified independently based on their

phenotype in cell culture and have a transposon inserted at

different position of the same gene, while deriving from the same

PG2 type strain.

The almost strict conservation of the NIF locus among all

mycoplasma genomes sequenced so far emphasizes its biological

importance (Table S1), and raises questions regarding the role of

this particular locus during the infectious process. As expected, the

highest level of similarity was observed for the closely related M.

bovis, suggesting that CA in small ruminants can provide an

interesting model to study M. bovis infection in cattle. In M.

agalactiae, the NIF locus consists of two CDSs encoding

homologues of nitrogen fixation proteins NifS and NifU, two

proteins involved in iron-sulfur Fe-S cluster biogenesis. Fe-S

clusters are ubiquitous and versatile prosthetic groups found in

enzymes catalyzing a variety of redox reactions or proteins serving

as redox sensors in a broad range of regulatory processes [23].

Bacteria use three distinct systems for Fe-S cluster biogenesis: ISC,

SUF, and NIF machineries. Initially annotated as NIF homo-

logues, sequence features of the SUF machinery were clearly

identified in M. agalactiae NifS and NifU proteins that were

Figure 5. Stochastic changes in M. agalactiae antigenic structure upon colonization of the animal host. (A) SDS-PAGE analysis of Triton X-114 soluble proteins extracted from cultures of M. agalactiae recovered from milk samples collected from animals A5 and A8 at different days post-inoculation (0 to 28). (B) Characterization of the Vpma expression profile using specific polyclonal antibodies (see Materials and Methods). Alipoprotein P80 antiserum was used as control. Molecular weight standards (MW) are in kDa.doi:10.1371/journal.pone.0093970.g005

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 10 April 2014 | Volume 9 | Issue 4 | e93970

indicative of a cysteine desulfurase (SufS) and a protein scaffold

(SufU), respectively [18]. In contrast to many bacteria, SUF is the

unique Fe-S biogenesis system in most Firmicutes and in mycobac-

teria [20,24]. Interestingly, SUF is associated with stress tolerance

in many bacteria by ensuring the assembly of Fe-S proteins under

conditions of iron limitation and oxidative stress, and has been also

linked to virulence in a few pathogenic species [25,26]. Therefore,

it is tempting to speculate that these two proteins might play an

essential role in M. agalactiae survival within the host by their

contribution in the biosynthesis of iron-sulfur clusters. However,

given the multiple biological functions involving cysteine desul-

furases and scaffold proteins [27], further studies are needed to

define precisely the contribution of the NIF locus in the infectious

process of M. agalactiae. The absence of an IgG response in animals

inoculated with the NifS mutants suggests a critical role of the NIF

locus in the initial steps of colonization by the mycoplasma. This

hypothesis is consistent with the rapid loss of viability of the two

NifS mutants upon co-culture with mammalian cells [18].

Proteins involved in metabolic regulation or stress response in

bacteria can have additional biological functions and some are

now regarded as virulence factors [28]. This capacity of some

bacterial proteins to exhibit more than one function, a concept

known as protein moonlighting, is highly appealing in the context

of the reduced mycoplasma coding capacity. Studies with several

mycoplasma species, including the human pathogen M. pneumoniae,

have identified moonlighting functions in several glycolytic

enzymes and chaperones including the glyceraldehydes-3-phospa-

hate dehydrogenase, the pyruvate dehydrogenase subunit B,

proteins with enolase activity, and the elongation factor Tu [28–

33]. The moonlighting function associated with these proteins is

conditioned by their location at the mycoplasma surface [28].

Whether multifunctional enzymes, such as proteins encoded by

the NIF locus, may have a moonlighting activity in mycoplasmas

remains highly speculative, but our study clearly demonstrate a

role for housekeeping genes in M. agalactiae virulence.

While experimental models of infection with M. agalactiae were

developed using elevated doses that ranged from 105 to 109 CFUs

[34–38], the model validated in the present study suggests that

even minimal infectious doses, as low as 103 CFU, can establish

infection in lactating ewes and disseminate to various body sites

including lymph nodes and the mammary gland. This greatly

expands the number of mutants that can be tested simultaneously

in a single animal, and provide a useful system for in vivo functional

studies using large pools of mutants in a signature-tagged

mutagenesis like strategy.

A major drawback to performing genetic screens in cell culture

is the absence of host immune defenses. To secure their survival,

mycoplasmas have evolved sophisticated mechanisms of phase and

antigenic variation [12]. Phase variation have been shown to occur

in vitro with a high frequency (1022 to 1025 events/cell/

generation) resulting in highly heterogeneous populations. It is

assumed that high-frequency phase and size variation of surface

components in Mollicutes that colonize immunocompetent hosts

largely contributes to escaping the host immune response. Our

results confirmed the rapid modification of M. agalactiae surface

architecture upon replication in vivo, which can occur even after a

single passage in the animal host, and identified Vpmas surface

lipoproteins as major contributors of antigenic variation in the

natural host. The remarkable plasticity of M. agalactiae was further

documented by the diversity of the antigenic profiles identified that

differed considerably, not only between animals but also between

samples in the same animal, with different profiles coexisting at the

same time. M. agalactiae Vpma profile in udder-halves of animals

exhibiting a continuous excretion profile was characterized by

important shifts in expression, providing the first detailed

description of sequential changes affecting the surface architecture

of this pathogen during the course of infection in its natural host. It

would be interesting to determine if these shifts are consecutive to

a re-infection of these animals by antigenic variants excreted by

other animals in the group. This may have important implications

for the circulation and persistence of M. agalactiae within the herd.

The role of surface antigenic variation in the pathogenesis of

mycoplasma infections is a recurrent question. Infection studies

with the avian pathogen M. gallisepticum showed that antigenic

variation is detectable within a few days after infection, before the

onset of a serological response [39]. While surface antigenic

variation may facilitate infection by M. gallisepticum, a recent study

with M. agalactiae phase-locked mutants expressing only a single

Vpma surface lipoprotein suggested that Vpma phase variation

was not necessary for establishing an infection in the ovine host,

but might critically influence the survival and persistence of the

pathogen under natural field conditions [38]. In the present study,

PG2 infected animals exhibited marked differences in the kinetic

of mycoplasma excretion in milk and the development of the

serological response. While all PG2 inoculated animals developed

specific IgG antibodies, the onset of the response varied

considerably among animals. Although the number of animals

was low, it is noteworthy that animals A4, A5 and A8, which were

all characterized by a prolonged mycoplasma excretion, developed

a serological response only after the beginning of the excretion,

while animals that were characterized by a single excretion day

exhibited a specific IgG response prior to mycoplasma detection in

milk samples. Further studies involving larger experimental groups

are needed to evaluate the influence of the serological response

against M. agalactiae on the duration of mycoplasma excretion in

the ovine host. This result, together with the rapid diversification

of Vpma expression profile upon M. agalactiae replication in the

animal host, raise important questions regarding the influence of

these variations on functional studies involving infections with

knock-out mutants. These results underscore the importance of

characterizing the Vpma profile of knock-out mutants.

The avirulent phenotype exhibited by the NifS mutants suggests

that key functions contributing to M. agalactiae survival and

dissemination in the ovine host can be identified upon genetic

screening in cell culture, providing a convenient system for high-

throughput analysis of large transposon mutant libraries and the

study of virulence factors in a large number of pathogenic

mycoplasma species. Host-colonization by mycoplasmas is a

valuable criterion to compare the infectivity of knock-out mutants

in vivo, a way to overcome some limitations linked to individual

susceptibility and to combine genetic screens in cell culture with

in vivo studies. Finally, M. agalactiae replicating in the ovine host

may also provide valuable information regarding Mycoplasma bovis-

host interaction, given the genetic proximity of these two

pathogenic species.

Supporting Information

Table S1 Conservation of M. agalactiae NifS among main

pathogenic mycoplasma species.

(DOC)

Acknowledgments

We thank Aude Bressolin and Anne-Claire Seguret for excellent technical

assistance. We also thank Agnes Skapski for fruitful discussions.

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 11 April 2014 | Volume 9 | Issue 4 | e93970

Author Contributions

Conceived and designed the experiments: EB DB XB CC. Performed the

experiments: EB DB ES MCH PR XB CC. Analyzed the data: EB CC.

Contributed reagents/materials/analysis tools: EB CC. Wrote the paper:

EB CC.

References

1. Razin S, Yogev D, Naot Y (1998) Molecular biology and pathogenicity of

mycoplasmas. Microbiol Mol Biol Rev 62: 1094–1156.2. Sirand-Pugnet P, Citti C, Barre A, Blanchard A (2007) Evolution of mollicutes:

down a bumpy road with twists and turns. Res Microbiol 158: 754–766.3. Razin S, Hayflick L (2010) Highlights of mycoplasma research - an historical

perspective. Biologicals 38: 183–190.

4. Hutchison CA, Peterson SN, Gill SR, Cline RT, White O, et al. (1999) Globaltransposon mutagenesis and a minimal Mycoplasma genome. Science 286:

2165–2169.5. Glass JI, Assad-Garcia N, Alperovich N, Yooseph S, Lewis MR, et al. (2006)

Essential genes of a minimal bacterium. Proc Natl Acad Sci U S A 103: 425–

430.6. Gibson DG, Glass JI, Lartigue C, Noskov VN, Chuang RY, et al. (2010)

Creation of a bacterial cell controlled by a chemically synthesized genome.Science 329: 52–56.

7. Bergonier D, Berthelot X, Poumarat F (1997) Contagious agalactia of smallruminants: current knowledge concerning epidemiology, diagnosis and control.

Rev Sci Tech Off Int Epizoot 16: 848–873.

8. Corrales JC, Esnal A, De la Fe C, Sanchez A, Assuncao P, et al. (2007)Contagious agalactia in small ruminants. Small Rum Res 68: 154–166.

9. World Organisation for Animal Health (2012) OIE Manual of Diagnostic Testsand Vaccines for Terrestrial AnimalsSeventh Edition. OIE Biological Standards

Commission, Paris, France. Available: http://www.oie.int/en/international-

standard-setting/terrestrial-manual/access-online/.10. Sirand-Pugnet P, Lartigue C, Marenda M, Jacob D, Barre A, et al. (2007) Being

pathogenic, plastic, and sexual while living with a nearly minimal bacterialgenome. PLoS Genet 3: e75.

11. Nicholas RAJ (2011) Bovine mycoplasmosis: silent and deadly. Vet Rec 168:

459–462.12. Citti C, Nouvel LX, Baranowski E (2010) Phase and antigenic variation in

mycoplasmas. Future Microbiol 5: 1073–1085.13. Nouvel LX, Marenda M, Sirand-Pugnet P, Sagne E, Glew M, et al. (2009)

Occurrence, plasticity, and evolution of the vpma gene family, a genetic systemdevoted to high-frequency surface variation in Mycoplasma agalactiae. J Bacteriol

191: 4111–4121.

14. Cacciotto C, Addis MF, Coradduzza E, Carcangiu L, Nuvoli AM, et al. (2013)Mycoplasma agalactiae MAG_5040 is a Mg2+-dependent, sugar-nonspecific SNase

recognised by the host humoral response during natural infection. PLoS ONE 8:e57775.

15. Chopra-Dewasthaly R, Zimmermann M, Rosengarten R, Citti C (2005) First

steps towards the genetic manipulation of Mycoplasma agalactiae and Mycoplasma

bovis using the transposon Tn4001mod. Int J Med Microbiol 294: 447–453.

16. Zimmerman CU, Herrmann R (2005) Synthesis of a small, cysteine-rich, 29amino acids long peptide in Mycoplasma pneumoniae. FEMS Microbiol Lett 253:

315–321.17. Halbedel S, Stulke J (2007) Tools for the genetic analysis of Mycoplasma.

Int J Med Microbiol 297: 37–44.

18. Baranowski E, Guiral S, Sagne E, Skapski A, Citti C (2010) Critical role ofdispensable genes in Mycoplasma agalactiae interaction with mammalian cells.

Infect Immun 78: 1542–1551.19. Skapski A, Hygonenq MC, Sagne E, Guiral S, Citti C, et al. (2011) Genome-

scale analysis of Mycoplasma agalactiae loci involved in interaction with host cells.

PLoS ONE 6: e25291.20. Fontecave M, Ollagnier de Choudens S, Py B, Barras F (2005) Mechanisms of

iron-sulfur cluster assembly: the SUF machinery. J Biol Inorg Chem 10: 713–721.

21. Marenda MS, Sagne E, Poumarat F, Citti C (2005) Suppression subtractive

hybridization as a basis to assess Mycoplasma agalactiae and Mycoplasma bovis

genomic diversity and species-specific sequences. Microbiology 151: 475–489.

22. Chopra-Dewasthaly R, Citti C, Glew MD, Zimmermann M, Rosengarten R,et al. (2008) Phase-locked mutants of Mycoplasma agalactiae: defining the molecular

switch of high-frequency Vpma antigenic variation. Mol Microbiol 67: 1196–

1210.23. Beinert H, Holm RH, Munck E (1997) Iron-sulfur clusters: nature’s modular,

multipurpose structures. Science 277: 653–659.24. Py B, Barras F (2010) Building Fe-S proteins: bacterial strategies. Nat Rev

Microbiol 8: 436–446.

25. Huet G, Daffe M, Saves I (2005) Identification of the Mycobacterium tuberculosis

SUF machinery as the exclusive mycobacterial system of (Fe-S) cluster assembly:

evidence for its implication in the pathogen’s survival. J Bacteriol 187: 6137–6146.

26. Runyen-Janecky L, Daugherty A, Lloyd B, Wellington C, Eskandarian H, et al.(2008) Role and regulation of iron-sulfur cluster biosynthesis genes in Shigella

flexneri virulence. Infect Immun 76: 1083–1092.

27. Hidese R, Mihara H, Esaki N (2011) Bacterial cysteine desulfurases: versatile keyplayers in biosynthetic pathways of sulfur-containing biofactors. Appl Microbiol

Biotechnol 91: 47–61.28. Henderson B, Martin A (2011) Bacterial virulence in the moonlight: multitasking

bacterial moonlighting proteins are virulence determinants in infectious disease.

Infect Immun 79: 3476–3491.29. Dallo SF, Kannan TR, Blaylock MW, Baseman JB (2002) Elongation factor Tu

and E1 beta subunit of pyruvate dehydrogenase complex act as fibronectinbinding proteins in Mycoplasma pneumoniae. Mol Microbiol 46: 1041–1051.

30. Balasubramanian S, Kannan TR, Baseman JB (2008) The surface-exposed

carboxyl region of Mycoplasma pneumoniae elongation factor Tu interacts withfibronectin. Infect Immun 76: 3116–3123.

31. Chen H, Yu S, Shen X, Chen D, Qiu X, et al. (2011) The Mycoplasma gallisepticum

a-enolase is cell surface-exposed and mediates adherence by binding to chicken

plasminogen. Microb Pathog 51: 285–290.32. Dumke R, Hausner M, Jacobs E (2011) Role of Mycoplasma pneumoniae

glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in mediating interactions

with the human extracellular matrix. Microbiology 157: 2328–2338.33. Thomas C, Jacobs E, Dumke R (2013) Characterization of pyruvate

dehydrogenase subunit B and enolase as plasminogen binding proteins inMycoplasma pneumoniae. Microbiology 159: 352–365.

34. MacOwan KJ, Brand TF, McGillveray N, Hunter AR (1984) Experimental

infection of castrated lambs with Mycoplasma agalactiae. J Hyg (Lond) 93: 455–463.35. Sanchis R, Abadie G, Lambert M, Cabasse E, Guibert JM, et al. (1998)

Experimental conjunctival-route infection with Mycoplasma agalactiae in lambs.Small Rum Res 27: 31–39.

36. Buonavoglia D, Fasanella A, Greco G, Pratelli A (1999) A study of anexperimental infection of sheep with Mycoplasma agalactiae. New Microbiol 22:

27–30.

37. Sanchis R, Abadie G, Lambert M, Cabasse E, Dufour P, et al. (2000)Inoculation of lactating ewes by the intramammary route with Mycoplasma

agalactiae: comparative pathogenicity of six field strains. Vet Res 31: 329–337.38. Chopra-Dewasthaly R, Baumgartner M, Gamper E, Innerebner C, Zimmer-

mann M, et al. (2012) Role of Vpma phase variation in Mycoplasma agalactiae

pathogenesis. FEMS Immunol Med Microbiol 66: 307–322.39. Glew MD, Browning GF, Markham PF, Walker ID (2000) pMGA phenotypic

variation in Mycoplasma gallisepticum occurs in vivo and is mediated bytrinucleotide repeat length variation. Infect Immun 68: 6027–6033.

Virulence Factor of Mycoplasma agalactiae

PLOS ONE | www.plosone.org 12 April 2014 | Volume 9 | Issue 4 | e93970